This Content Is Only For Subscribers
It is not enough to develop and administer a good test or assignment and score it without using a markscheme. A markscheme may be called by different names such as marking guides, marking criteria, marking matrices, rubrics, marking keys. But what is important is that a markscheme is a tool developed for use to ensure fairness in scoring assessment of learners’ work. Scoring is said to be fair when different markers, award the same mark to the same work and that the resulting marks effectively discriminate between learners.
In a well-constructed test, marks are shown at the end of every item. These marks are in a way a suggestion of how much each question weighs/contributes to the overall test. In a multiple-choice item test for example, all items weigh the same, while in a structured items or essay test, the marks/weights are different. Thus scoring structured items requires the use of a standardised criteria to promote fairness and objectivity.
Scoring without using a marksheme, may results in some learners being awarded undeserving marks or denied deserving marks. Once a markscheme has been developed, anybody who has ample background in the subject can mark test/assignment fairly and produce the same score consistently.
It is not always easy to develop a marking scheme. However, multiple choice items or objective items marksheme is the simplest to develop and use, because the answer is either correct or incorrect. For other structured and subjective items such open-ended items essays, the correctness of the answer lies in a continuum. Therefore, the marking scheme should help award the marks accordingly.
The development of a markscheme requires teamwork and thorough reviewing. Even after reviewing, it has to be piloted by using it to score some few learners tests or assignments. Learners responses best refines the marking scheme and validates it. As the scoring continues, leaners new responses that warrant to be credited will keep coming up and they will have to be added. It must be stressed that a mark scheme is a working document, in many cases further developed and expanded on the basis of learners’ reactions to a particular item.
Depending on the items in a test, markschemes can be developed as analytical or holistic rubrics. A rubric is the combination of the scoring criteria with a rating scale. Thus rubrics define expectations for learner on a specific task and helps to ensure consistency in the evaluation from learner to learner.
When marking is done using holistic rubric, the scorer scores the overall process or product as a whole without judging the component parts separately and a single mark is given for the piece of work as a whole. For example, in D&T subject, learners can be holistically scored on, say a chair they have made in the workshop, by looking at the overall quality of the chair. Thus holistic scoring is easier to do but does not provide detailed information on learner performance for each criteria.
On the other hand, when scoring using analytic rubric, marks are given for individual criteria in the marking scheme, using different descriptive ratings. Each criterion receives a separate score. For example, when the chair is scored analytically, the scorer scores the quality of the joints, the quality of gluing, the quality of varnishing, and so on. Analytic rubric scoring produces more reliable marking with greater agreement between markers.
Scoring using analytical rubrics take more time to score but provides more detailed image of learner performance. This results in improved learning and achievement. Part of the reason there is little improvement in learning is because majority of those who score learners work use holistic rubrics or in acute situations do not use markschemes at all.